基于项目反应理论构建睡眠ICF综合核心组合精要模型的研究

关键词：国际功能、残疾和健康分类  项目反应理论  康复  睡眠

Keywords:international classification of functioning  disability and health  item response theory  rehabilitation  sleep disorder

摘要：

      摘要 目的：基于项目反应理论（item response theory， IRT）构建具备评估效能的精要版睡眠障碍ICF综合核心组合（parsimonious sleep disorder ICF comprehensive core sets），并为评估与治疗衔接提供依据。 方法：基于最大变异策略抽样原则采集便利样本，最大变异的控制变量包括年龄、体质指数（body mass index， BMI）、健康测量12项简表（the 12-item short form health survey， SF-12）的生理总评分（Physical component summary， PCS）和心理总评分（mental component summary， MCS）、匹兹堡睡眠质量指数（pittsburgh sleep quality index， PSQI）、症状自评量表—90（the symptom checklist 90， SCL-90）。对符合纳排标准的160例睡眠障碍患者进行睡眠障碍ICF综合核心组合、PSQI、蒙特利尔认知评估（Montreal cognitive assessment， MoCA）、SCL-90、SF-12等问卷调查。通过IRT构建精要版睡眠障碍ICF综合核心组合，使用蒙特卡洛模拟得到人项图分析个体睡眠相关功能水平（θ）值与类目难度之间的关系，计算匹配条目的潜在获益度。 结果：构建了包含34个类目的3参数逻辑斯蒂模型（3-parameter logistic model， 3PLM），该模型具有较高的信度（内部一致性信度检验结果：Cronbach α=0.9249，Guttman's λ-2=0.9320，MS统计=0.9398，LCRC=0.9585）和效标效度（34项ICF类目得分与ICF原始评分、PSQI评分、SCL-90及SF-12评分均有显著相关，P＜0.001）。本研究采用人项图对个体睡眠相关功能水平（θ）值与类目难度进行匹配分析，区分出预期胜任和预期挑战类目，并计算了每个匹配条目的潜在获益度。类目难度与获益度呈强相关（P<0.001，Pearson相关系数为0.777）。 结论：本研究基于IRT构建了包含34个类目精要版睡眠障碍 ICF 综合核心组合。该模型可以有效评价睡眠障碍人群睡眠相关（包括认知行为、心理情绪、社会）功能水平。模型应用分析进一步提供了可视化工具和干预类目后的预期获益度，可以为睡眠障碍康复提供新思路。

Abstract:

      Abstract Objective： To develop an essential version of the Sleep Disorder ICF combination with assessment efficacy based on item response theory （IRT）， and provide a basis for the connection between assessment and treatment. Method： Based on the sampling principle of the maximum variation strategy， convenience samples were collected. The control variables of maximum variation included age， body mass index （BMI）， the physical component summary （PCS） and mental component summary （MCS） of SF-12， the Pittsburgh Sleep Quality Index （PSQI）， and the Symptom Checklist 90 （SCL-90）. ICF core set of sleep disorders， Pittsburgh Sleep Quality Index （PSQI）， Montreal Cognitive Assessment （MOCA）， Symptom Checklist-90 （SCL-90）， and the 12-item Short Form Health Survey （SF-12） were administered to eligible participants .An abridged version of the ICF-based comprehensive core set was conducted for sleep disorders using IRT. Monte Carlo simulation was applied to analyze the relationship between individual functional levels （θ） and item difficulty in person-item maps， and to calculate the degree of benefit from interventions. Result：A 3-parameter logistic model （3PLM） with 34 categories was developed，demonstrating high reliability （internal consistency reliability test results： Cronbach's α=0.9249， Guttman's λ-2=0.9320， MS statistic=0.9398，LCRC=0.9585） and validity （the scores of the 34 ICF categories were significantly correlated with the original ICF scores， PSQI scores， SCL-90 scores， and SF-12 scores，P<0.001）. The person-item map was used to match individual sleep-related functional levels （θ） to the item difficulties， distinguishing between the mastered and challenging categories，The degree of benefit was calculated for each matched item, revealing a strong correlation between item difficulty and benefit degree（P<0.001， Pearson correlation coefficient was 0.777）. Conclusion： This study constructed a parsimonious version of the ICF-based comprehensive core set for sleep disorders with 34 items using IRT. The model can effectively evaluate the sleep-related （including cognitive-behavioral， psychological-emotional， and social） functional levels in individuals with sleep disorders. The application of the model analysis further provides a visual tool and the expected benefit after intervention items， which can offer new insights for the rehabilitation of sleep disorders.